
In the 22nd century visionary scientist Vannevar Morgan conceives the most grandiose engineering project of all time, and one which will revolutionize the future of humankind in space: a Space Elevator, 36,000 kilometers high, anchored to an equatorial island in the Indian Ocean.
Your comment reminds me of a funny aside in Arthur C Clarke’s The Fountains of Paradise (wonderful novel, about the construction of a space elevator). A famous architect, designing the world’s longest and highest bridge, fights to avoid putting guard rails along the sides. His justification is a) cars are all computer-controlled and totally reliable in the book’s setting, so rails are unnecessary; some suspect it’s actually b) the bridge will look nicer without guard rails; or even darker, c) if a car does somehow go off-course, he’d prefer it doesn’t damage the bridge before plummeting half a mile into the sea.
Arthur C Clarke wrote a great story about space elevators called "The Fountains of Paradise".
Let's see ... at the time he wrote "Neuromancer" William Gibson had never used a computer -- he wrote it on a manual typewriter (and used the royalties to buy an Apple IIc). His creative vision was more one of social atomization and a particular aesthetic in a multi-national dominated future than it was informed by actual technology. Despite which, his creative vision of "cyberspace" got taken seriously enough as a visual metaphor that I remember sitting in on a debate at the W3C conference in 1996 between folks proposing that VRML should follow a Gibsonian model, and those who thought "Snow Crash" was the way forward. (Neal Stephenson's followers won the day. See also "Second Life", I guess ...)
That's an example of "influencing". "Neuromancer" was very influential, but hardly predictive.
Actual predictive SF is much, much rarer, but if we ever get a space elevator you can probably blame Arthur C. Clarke ("The Fountains of Paradise"), which is an old-school didactic hey-why-don't-we-do-this-the-numbers-check-out work of prediction in fiction drag. Or maybe Andy Weir's "The Martian".
Written SF has traditionally been utter rubbish at predicting developments in the IT/computing sector, with notable exceptions -- for example the novella "True Names" by Vernor Vinge. (I've attempted this too but I'm not going to bang my own drum here.)
So dune is a great book but it is 90% fantasy, there is only a little bit of research done by the author on desert ecosystems.
My favourite real science fiction author is Arthur C clarke. some good books by him...
earthlight
the fountains of paradise
islands in the sky
the sands of mars
rama
songs of distant earth
Also would recommend accelerando by Charles Stross and the mars series by Kim Stanley Robinson
